arXiv:astro-ph/0703507v1 19 Mar 2007

Quark core formation in spinning-down pulsars.

G. F. Marranghello, T. Regimbau, J. A. de Freitas Pacheco
February 5, 2008
Observatoire de la Cote dAzur, Nice - France
Abstract
Pulsars spin-down due to magnetic torque reducing its radius and increasing the central energy density. Some pulsar which are born with
central densities close to the critical value of quark deconfinement may
undergo a phase transition and structural re-arrengement. This process
may excite oscillation modes and emmit gravitational waves. We determine the rate of quark core formation in neutron stars using a realistic
population synthesis code.

Introduction

Black holes (BHs) and neutron stars (N Ss) are certainly two major potential sources of gravitational waves (GW s). Unlike BHs, whose gravitational
waveforms are specified essentially by their masses and angular momenta, the
characteristics of the gravitational emission from N Ss depend on the properties
of the nuclear matter.
Different mechanisms related to N Ss susceptible to produce large amounts of
GW s have been investigated in the past years (see de Freitas Pacheco 2001 for a
recent review). In particular, the mini gravitational collapse induced by a phase
transition in the core[1]. If quark deconfinement occurs in the central region
of the N S, the core will have a softer equation of state, inducing the system
to search for a new equilibrium configuration, which will be more compact and
having a larger binding energy. The energy difference will partially cover the cost
of the phase transition and will be partially used to excite mechanical modes,
which will be damped either by heat dissipation or gravitational wave emission.
Radial oscillation modes are more likely to be excited after the mini-collapse if
the star has a slow rotation. In this case, most of the mechanical energy will be
dissipated in the form of heat and radiated away. If the N S has an important
rotation when the conditions for deconfinement are attained, then non-radial
oscillations and radial modes coupled to rotation[2] may lead to an important
GW emission, whose energy amounts to about 105253 erg.
For a given equation of state for the hadronic and for the quark matter,
the deconfinement occurs when the Gibbs conditions are satisfied, e.g., equality
1

between pressure and chemical potential of both phases. If the baryonic mass
of the configuration is high enough, the pressure and the energy density in the
central regions attain values required for a phase transition to occur. However,
if the star is in rapid rotation, the central pressure and energy density are below
the critical values and the N S has an internal structure constituted essentially
by hadrons. If, as expected, the N S has a magnetic field, the rotation velocity
will decrease due to the canonical magnetic dipole braking mechanism. Thus,
after a certain time, the phase transition conditions are reached and the star
develops a quark core.
If one assumes that the above scenario is correct, then a natural question
appears. What is the expected frequency of these mini-collapse events ? In
a recent essay [3] based on very simple arguments and supposing that Soft
Gamma Repeaters or cosmological Gamma-Ray Bursts are a consequence of
N Ss which have underwent a quark-hadron phase transition, estimated a frequency of 105 yr1 per galaxy for these events. In the present work, a more
detailed estimate of the occurrence of these events is given. Rotating N S models have been computed for the equation of state derived by MVP02 in order
to estimated, for a given baryonic mass, what is the critical rotation velocity
below which the phase transition is possible. Then, using numerical simulations
as in [4], the N S flux (in the P -P plane) crossing the critical region where the
transition occurs was estimated and, as a consequence, the event frequency.
Using the RNS rotation code[5] we are able to compute the evolution track of
neutron stars with constant baryonic masses. Stars with baryonic masses lower
then 1.05 M , which represent 10% of the whole population using a gaussian
distribution centered in the 1.4 M , do never form quark cores due to its low
central pressure. In opposite, 70% of the stars are already born with quark
cores due o its high central pressure. The figure 1 describes the stars with
masses between these two values and the formation of a quark core in a precise
rotation frequency.

The phase transition frequency

As we have seen in the previous section, only N Ss born in the mass interval
1.05 < M < 1.26 will develop a deconfined core. For a N S of a given mass
within that range, if its rotation velocity is not zero, the central density will
be below the critical value required for the phase transition to occur. During
its evolution the rotation frequency decreases due to magnetic torques and the
central density will eventually reaches the transition point. The timescale for the
occurrence of the transition depends on the initial rotation period and magnetic
field. The equation governing the number of N S in the rotation period P space
is
N (P, t) (P N (P, t))
+
= S(P, t)
(1)
t
P

c (10 g/cm )

7.0

14

6.0

5.0

4.0
0.0

1000.0

2000.0

3000.0

4000.0

5000.0

(Hz)

Figure 1: Evolutionary tracks for neutron stars of different masses in the diagram 0 . Rotation periods evolve according to the magnetic dipole braking
mechanism. In the upper shadowed region stars are already born with a quark
core whereas in the lower shadowed region, stars will never develop a quark core.
The region between the two horizontal lines corresponds to the energy density
jump during the phase transition.
where N (P, t) is the number of N Ss at the instant t with period P in the interval
P (P + dP ), S(P, t) is the source function and P dP/dt = f (P, H) describes
the deceleration mechanism. Formally, in terms of the Greens function, the
solution of this equation can be written
Z t
Z
N (P, t) =
dto dPo S(Po , t)G(P Po , t to )
(2)

which depends on the initial distribution of periods (Po ) and magnetic fields
defining the evolution rate dP/dt. It worth mentioning that, as consequence of
this relation, the period P is unambiguously connected with the initial period
Po at to .
In the present work, a different approach was adopted. We have assumed
that the initial distribution of rotation periods and magnetic fields are independent on the NS mass, which obeys a Gaussian distribution as we have already
mentioned. Under these conditions, the number of evolving N Ss that cross the
interval P (P + dP ) is given by the current J (Phinney & Blandford 1981)
1 X dP
(3)
P
dt
The current J was calculated using the population synthesis code developed
J=

by Regimbau & de Freitas Pacheco(2001), up-graded to take into account new

pulsars discovered at high frequencies by the Parkes Multibeam Survey and more
recent models for the natal kick distribution ([7]). Pulsars are generated at a
constant rate using a Monte Carlo procedure and their evolution in the Galaxy
is followed during hundreds of million years. An interactive method permits
modifications of the initial distributions until observed distributions like those
of the period, period derivative, magnetic field, distances are well reproduced
when selection effects are taken into account (for details, the reader is referred to
[4]). The derived parameters defining the properties of pulsars at birth are given
in table 4. The estimated number of active pulsars in the Galaxy is about
250,000, their birth rate is about one every 90 years and their mean lifetime is
about 22 million years.

Table 1: Parameters of the initial period P0 and magnetic braking timescale ln

0 distribution, assumed to be Gaussian
mean
dispersion
P0 (ms) = 240 20
P0 = 80 20
ln0 (s) = 11 0.5 ln0 = 3.6 0.2

In figure 2 is shown the calculated current J within bins of P (P + P ).

These bins are equivalent to mass bins M in the interval Mmin Msup and
the periods correspond to critical values for which the transition occurs, which
depends on the N S mass. The sum of the current in each bin, weighted by the
number of pulsars in that mass interval and averaged over the pulsar lifetime
gives a phase transition frequency of 1.2 105 events per year.

Gravitational Waves

Once we have determined the maximum distance probed (see [1]) by gravitational wave detector and the rate of events in our galaxy, we assume that
extragalactic pulsars are formed with the same characteristics of those found in
our galaxy. We have extrapolated our results to extragalactic calculations using
the data found in ([9, 8]) from where we extracted the luminosity of close galaxy
and clusters. We have used the luminosity as an indication of the formation of
stars and consequently of the formation of NS. The luminosity of the enclosed
members was divided by the Milky Way luminosity to derive the total number of detectable events. From ([9, 8]) we can extract, for example, the Local
Group of Galaxies, where our Galaxy is the most important, followed by M31
4

15.0

Log10 J +25

10.0

5.0

0.0 4
10

10

10

10

P (s)

Figure 2: We show the neutron stars flow by second for each period and the
flow of hybrid star formation in the filled bars.
and M33. We can also find at larger distances, objects like NGC672 (D6 Mpc
and L 6.7 1010 L ). Using this results we estimate the maximum detection
rate of only one event each 800 years for the present Virgo planned sensitivity,
which is low rate of events. This value get better by a factor of 2 if we are able
to detect event up to 10 Mpc (Adv. Ligo planned sensitivity) and reach one
event each 100 years if the detectors are able to see this kind of events in the
Virgo cluster.

Conclusions

Neutron stars evolution has been described. They were generated with periods
and masses that reproduces the characteristics of the detected pulsars. Than,
we have performed the spin-down due to magnetic torque, which increases the
central density of the stars. Some of these pulsars, borned with central densities
close to the deconfinement density, may undergo a phase transition and suffer
a micro-collapse. The rate such events were determined as well as the possible
rate of GW detection that come from them.
Most of the neutron stars cores are formed during the supernovae event.
The stars which form a quark core after the supernovae explosion have their
evolution determined by the initial mass and period. Our model has predicted
a rate of 105 events year1 galaxy 1 . As these events can be detected by the
gravitational wave detectors for distances close to 7 Mpc (13 Mpc) by the Virgo
(Adv. Ligo) detectors, we can, in first approximation, extrapolate the results
obtained from our galaxy to a greater distance which encloses more events and
5

Log10 (Number of Pulsars)

2.0

0.0

2.0

4.0
3.0

2.0
Log10 (P)

1.0

Figure 3: The number of pulsars which shall form a quark core as a function of
period in our galaxy for a simulated population about 250.000 pulsars with an
active life of 22106yrs.
estimate a rate of detection of about one each 800 years for the Virgo detector.
The planned advanced Ligo detector shall be able to see such events close to the
Virgo cluster and possibly detect one event each 100 years. This rate of events
can still get better if one takes into account the stars that undergo a phase
transition due to mass accretion, however we do not expect a great change in
the actual value since there are only few low-mass stars that could undergo the
phase transition.

References
[1] Marranghello G. F., Vasconcellos C. A. Z., and de Freitas Pacheco J. A.
(2002) PRD 66, 064027.
[2] Chau W.-Y. (1967) ApJ 147, 664.
[3] Ma F. (2002) Gen. Rel. Grav. 34, 1319.
[4] Regimbau T. and de Freitas Pacheco J. A. (2001) A&A 376, 381.
[5] Stergioulas N. (2003) Living Reviews in Relativity lrr-2003-3.
[6] Phinney E. S., Blandford R. D. (1981) MNRAS, 194, 137.
[7] Arzoumanian Z., Chernoff D.F. and Cordes J.M., 2001, ApJ, 568, 289.

[8] Niklas S., Klein U., Braine J. and Wielebinski R. (1995) A&AS 114, 21.
[9] Gourgoulhon, E., Chamaraux P. and Fouque P. (1992) A&A 255, 69.